GDC is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

View, browse and sort the ever-growing list of sessions by pass type, topic, and format. With this Session Scheduler, you can build your schedule in advance and access it during the show via export or with the GDC Mobile App, once live. Sessions do fill up so please arrive early to sessions that you would like to attend. Adding a session to your schedule does not guarantee you a seat.

Machine Learning Summit: Avoid Mistakes When Using Deep Learning for Testing and NPCs

Jeffrey Shih (Lead Product Manager AI, Unity Technologies)

Marwan Mattar (Senior Engineering & Research Manager, Unity Technologies)

Robin Nillson (Co-Founder, Carry Castle)

Pass Type: All Access Pass, Core+Summits Pass, Summits Pass - Get your pass now!

Topic: Programming

Format: Session

Vault Recording: TBD

Audience Level: All

There have been a lot of discussions (and promise) on how deep reinforcement and imitation learning can be used for scaling playtesting and NPC creation in games. Innovative AI institutions like DeepMind and OpenAI have beaten top players in Go, Starcraft II, and Dota 2, which has led to a flurry of publication and novel research that is making its way to the gaming industry. However, for many game studios today, the hope and dream of creating a truly intelligent AI is in sharp contrast with the cost and implementation realities.

In this presentation, you will hear lessons learned from small and indie studios who are leveraging deep reinforcement and imitation learning in order to gain an advantage. You will also hear what advantages these studios are achieving and learn best practices.

You will hear real-world examples from two studios who are implementing reinforcement and imitation learning for their titles.


Attendees will come away with the practicality of implementing reinforcement or imitation learning in their game. Specifically what it can be used for, the differences between RL and IL bots, and common challenges during implementation.

Intended Audience

Game developers, AI engineers, game designers, QA / test engineers, business owners.